Multiple Quality Characteristics Prediction and Parameter Optimization in Small-Scale Resistance Spot Welding

作者：Wan, Xiaodong^*; Wang, Yuanxun; Zhao, Dawei

来源：ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING, 2016, 41(5): 2011-2021.

DOI：10.1007/s13369-016-2061-2

• Small-scale resistance spot welding
• Multi-objective optimization
• Principal component analysis
• Grey relational analysis
• Multiple regression analysis
• Neural network

This study focuses on prediction and optimization of multiple quality characteristics in small-scale resistance spot welding of titanium alloy. Grey relational analysis was first conducted to roughly estimate the optimum welding parameters combination. Multiple regression analysis was then implemented for a local parameter optimization. Optimum welding parameters were determined by desirability function and multi-objective genetic algorithm approach separately. A back propagation neural network model was also performed to simulate relationship between welding parameters and single output of the first principal component. Performance of the particle swarm optimization was better than genetic algorithm in obtaining optimum welding parameters. The neural network-based model was very effective in global optimization. A good agreement could be found between experimental results and predicted weld quality characteristics. Weld quality could be found significantly improved with the proposed methods.

• 出版日期2016-5
• 单位华中科技大学